MICHELE ROMANELLI, ENRICO TARULLI ICRAM – Via di Casalotti, 300 – Rome (Italy)
MECHANIZATION OF FISHING OPERATIONS BY FIXED GEARS: REPORT ON TESTS PERFOMED BY TWO
“AUTOMATION SYSTEMS” FOR BOTTOM LONGLINES IN THE SOUTHERN ADRIATIC AND NORTHEASTERN IONIAN SEAS. Key – words: Mechanization, longline, epibathyal, catches, Mediterranean,
Italy, Fishery.
Summary
During years 1991 – 1993 tests were intermittently carried out along the Apulian coast and in the NE Ionian Sea to assess the fishing performances of two “mechanization systems” for bottom longlines when operating at different depths. However, most data pertain to the period late June – late September 1993 during which a more continuos series of well–handled fishing trials were perfomed.
Our data showed that one of the mechanization systems being experimented effectively baited 70% - 90% ca of the available hooks when hard and soft– flesh fish were respectively used. Fishing yelds mainly ranged from 5 to 10 individuals and from 3 to 7 kilograms foe each hundred baited hooks. These yelds were low when compared with the output of two fishing operations carried out by means of one standard commercial longline. Nevertheless, experimental yields are supposed to be economically viable because of the high numbers of hooks that well–trained crews are expected to set daily using the automation systems.
In terms of number of individuals catches from experimental longlines were mainly composed of Scorpenids, Mediterranean hakes, conger eels and Selachians. Several clusters of fishes were outlined from our fishings data.
Riassunto
parte dei risultati esposti nel presente lavoro si riferiscono però al periodo fine giugno – fine settembre 1993, poiché solo in questo lasso di tempo fu possibile effettuare prove di pesca in maniera sufficientemente continuativa. I dati ottenuti hanno evidenziato che sistemi d’automazione del tipo di quelli da noi sperimentati sono in grado di innescare in maniera efficace una frazione all’incirca pari al 70% - 90% degli ami offerti, in relazione alla minore o maggiore consistenza delle carni dei pesci utilizzati. Nella maggior parte delle cale le rese di pesca sono risultate comprese tra 5 e 10 individui e tra 3 e 7 kilogrammi di pesce per ogni centinaio di ami correttamente innescati.
Questi rendimenti sono risultati nettamente inferiori a quelli registrati in due prove di pesca svolte con un palangaro di tipo commerciale in una delle nostre aree di studio. I rendimenti rilevati nel corso delle nostre prove sperimentali sono stati tuttavia ritenuti economicamente validi, in considerazione dell’elevato numero di ami che equipaggi addestrati dovrebbero essere in grado di calare giornalmente con sistemi d’automazione del tipo di quelli in esame.
In termini numerici le catture ottenute con palangari sperimentali sono risultate composte prevalentemente da Scorpenidi, gronchi, naselli e Selaci. In base alla composizione delle catture sono stati individuati alcuni “clusters” di specie tra loro più frequentemente associate.
Introduction
The depleted state of most marine fish stocks (FAO, 1993) stresses the need for using more selective (i.e. minimising non – commercial catches) and less disruptive fishing methods.
In this view, great interest is given to the “automation systems” developed since the late Seventies to mechanise fishing operation by angling as well as bottom and surface longlines (CADE, 1981; ANONYMOUS, 1984; VON BRANDT, 1984). Nodaways such deck systems are largely employed throghout the world in fisheries aimed both at demersal and pelagic resources (e. g. BJORDAL, 1990a; CHOPIN, 1994).
increased level of mechanization could also force some of the smaller trawlers to turn to fixed gears, thus reducing the negative impact on onshore fishing grounds.
In view of such potential benefits, two automation systems were tested in a set of hauls intermittently performed in the Southern Adriatic and the north – eastern Ionian Seas during the years 1991 – 1993. The tests were mainly aimed at assessing operational performance of the “automation systems” on deep grounds (down to 400 m) exploited by local fishermen as well as at developing proper longlines for this kind of fishery.
A specialised fishery based on bottom and surface longlines has been long operating in the areas concerned with our study (MARANO et al., 1988; PIETRUCCI and ANTOLINI, 1990a), but in our case most tests were perfomed by a common trawler to see how the crew coped with the sophisticated strategy required for fishing by fixed gears.
In more recent times the EC (now EU) passed new fishing regulations (Rule 1626/94) that reduce to 7,000 metres the maximum quantity of longlines to be stored aboard, so little room is left to automation systems like those tested during our trials. Nevertheless, we feel that our results deserve full reporting, since restrictions could be mitigated in the future and little is know about longlining in the Mediterranean Sea, Moreover, few independent reports exist on the performances of the automation systems for bottom longlines (e. g. DAHM, 1986a and b).
Material and methods - Tests at sea
The experimental hauls concerning with this paper were mainly perfomed during June, July and September 1993 by a trawler (capacity: 36.00 GT; power: 280 HP; crew: 4 persons) for testing two automation systems for bottom longlines, respectively produced in USA and Germany (hereinafter called systems or models I and II). Shorter or less continuos sessions of trials were carried out during summer or autumn of years 1991 – 1992, so limited sets of data from these periods are also reported.
while the hooks were positioned on rails and driven to storage boxes. Systems I and II mainly differ in their baiters, since in the first case the baiter is a large pipe filled with chopped fish where the hooks get baited in their fast run to the sea, whilst in model II the longline is paid outboard more slowly (expulsion rate: 0.5 – 1.0 vs. 1.0 – 3.0 ms-1, all calculations were done without considering the speed of the boat) through a narrow pipe where each hook encounters single baits falling by gravity from a side reservoir.
In our tests, longlines of different structure were used (Table 1), the snoods were composed of monofilament and their lengths ranged from 0.90 to 1.80 m (shorter snoods had resulted unproductive during trials carried out in summer 1992), whilst hooks were either Mustad Sea Kirby 2330 n˚ 6 or Olympus 9403 n˚ 7/0 (whose length, depth and gape were ca 40-20-15 and 65-25-20 mm respectively; see SOUSA et al, 1999, for terminology). However, most trials were performed by one longline (LLF1 in Table 1) whose mainline was composed of polypropylene (density: 0.91 g/cm3) so the gear was at some distance from the sea floor, allowing to operate on uneven or silty grounds (VON BRANDT, 1984). Very similar to LLF1 was also the gear used for two commercial hauls (i.e. performed by a boat routinely fishing by longlines) which we monitored in summer and autumn 1995 (see hereinafter), the main difference between the gears being the smaller hooks of the commercial longline (LLCM in table 1) and that it had lead weights connected to the mainline by 1.0 - 1.5 m snoods. Morevoer, it should be stressed that only one third of all experimental hauls (i. e. those during which setting and/or retrieval operations were carried out by one of two automation suystems) the hooks were actually baited by the systems themselves because we preferred to do it by hand to increase the number of hooks effectively baited and gather more information on the fishing performance of the gears. As far as fishing tactics are concerned, untrawlable areas were searched as they often host distinct assemblages of commercial fishes because of the existence of multiple microhabitats as well as the lower level of exploitation. Moreover, in areas with sharp slopes the coexistence of different water bodies sometimes increases plankton production and sedimentation rates for the organic matter, allowing commercial resources to be locally more abundant (CSANADY et al., 1988; CARTES et al., 1994).
were performed by system I (average number of hooks: 468+151; baited hooks: 437+143) and 7 by model II (average number of hooks: 654+153 baited hooks: 387+95); the fishing days were 26 and 6 respectively for the former system and the latter one. These experimental hauls were performed in an area 25-40 km to the N/NE of the harbour of Mola di Bari (Apulia, SE Italy), in proximity to the local submarine canyon (FA B B R I a n d GALLIGNANI, 1972; GALLIGNANI, 1973) (fishing A area in fig. 1; hauls = 41), then on grounds 10-14 km to the east of Otranto (fishing B area; hauls = 3) and just to the SE of the Greek island of Fano (fishing C area; hauls = 4). All experimental hauls by system II were performed in the fishing A area which was also selected for both commercial hauls, each numbering about 2,000 hooks, we observed in August and November 1995.
All experimental hauls of summer 1993 lasted from dawn to late morning, the minimum time of permanence at sea of baited hooks (i. e. the time calculated without considering the duration of the setting and retrieving operations) ranged from 1.5 to 2.5 hours. As a rule, baits were 6 - 8 g chops of fresh or thawed sardines, Sardina pilchardus (Walbaum); on one occasion 10 - 12 g pieces of fresh mackerel, Scomber scombrus L., were used. In commercial hauls the hooks were baited by intact sardines.
Depth and position of longlines were registered by echosounder and Loran. Hauls were carried out in the 110-405 m depth range.
As a rule fishes caught during experimental hauls were identified, measured and weighed. Total lengths were measured, in individuals of different species, approximating them to the lower limit of 0.5 or 1.0 cm; weights were obtained by two steelyard balances so they are intrisecally imprecise although single animals were formally weighed to the 0.01 or 0.1 kg (see Table 4). When many low-priced fishes were obtained in a single haul, they were simply counted and weighed together. In order to calculate the yields of the experimental fishing operations single weights (i. e. the weight of each fish) were summed up, neglecting small discrepancies due to a different precision of measures.
In both of the commercial hauls the total numbers of fishes caught and their average weights were estimated by counting the boxes filled for each species and making reasonable guesses on the average size of the fishes stored.
- Data analysis
groups (110 - 150, 150 - 300 and 300 - 405 m), without considering the sediment composition and the benthic communities involved since such information was lacking in most cases. Bathymetric limits were therefore identified to separate the hauls performed in proximity to the 1l0m contour line (where the average gradient of the Apulian shelf moves from 0˚ 39’ to 0˚ 16’; GALLIGNANI , 1973), as well as to the outer edge of the shelf and on the upper sector of the continental slope. It is also worth noting that the 1l0m contour line separates more coastal silty areas from outer sectors covered with “relict sand” (FA B B R I and G A L L I G N A N I, 1972; GALLIGNANI, 1973).
The consistency of the three assemblages of hauls was then tested by Canonical Discriminant Analysis (CDA, MASSART and KAUFMAN, 1989), based on the species composition of the catch observed for each haul. Because of the low number of hauls performed by system II (N = 7), partly with bad weather and untested longlines, most data concerning models I and II were put together.
Since the longlines presumably floated at some distance from the sea bottom, we calculated the average “hanging ratio” (i.e. the ratio obtained by dividing the total length of the mainline by the minimum distance existing between the two outermost buoys) for each haul to check whether the gear had been correctly positioned. In turn, the hanging ratios obtained and simple geometrical calculations based on the distribution of lead weights along the mainline allowed to estimate the average height of volutes formed (Table 1). In comparing data from hauls with different numbers of hooks, we assumed that catches were proportional, for a given area and species, to the quantity of hooks baited each time, neglecting the influence due to somewhat different snood spacing (MURPHY, 1960; RICKER, 1975; SKUD and HAMLEY, 1978; LOKKEBORG and BJORDAL, 1992).
following formulas: WSDN = ((∑(xi-X)2n
i)/m(Y-1)) 0.5
WSDW = ((∑(ci-C)2n
i)/m(Y-1)) 0.5
where xiis the frequency of fishes caught in the ithhaul, ci is the quantity of
fish (as kilograms) caught in the same haul for each hundred baited hooks, when the comparison was in terms of weights), X and C are the weighed means of the same parameters, m is the average number of hooks in the set of Y hauls (ARMITAGE, 1977).
Comparison between commercial and experimental hauls was however based on logical concepts, because the great discrepancy in the numbers of hooks used for the two groups of hauls made formal tests unfeasible.
Results
a - Baiting tests and longline retrieval
In nine hauls by system I and seven ones by system II the hooks were baited mechanically, registering for each haul the numbers of hooks baited in the time unit and the fraction of them retaining the bait itself when they struck the sea surface.
In eight out of the nine hauls by system I the baits were pilchard chops (average number of hooks: 544+137). and a total frequency of 71.2% “correctly baited” hooks (i.e. that retained the fish chop when dipping in the water) was registered. In the ninth haul with mechanical baiting by system I mackerel chops were used, registering a total frequency of 90.0% hooks correctly baited (300 hooks offered to the baiter). In similar tests carried out during summer 1992 with two similar longlines (LLF4 and LLFS in Table 1), baiting frequencies approximately ranged from 50% to 60% of the hooks offered when the baits were pilchards (hauls = 3; average number of hooks = 906+100), whilst the total frequency of baited hooks rose to 84.1% for 5 hauls where chopped mackerels were used (average number of hooks = 676+27).
On the whole the baiting performances of system II (unreported data) were worse than those summarised for model I; however it must be stressed that our tests were part of the development process of such second baiter. The “utilisation rate” of fish pieces (i.e. the percentage of fish chops that actually baited the hooks) was estimated around 55.0% for the baiter of system I fed with pilchard baits. About 0.9 Kg of chopped sardines, approximately corresponding to 130 fish pieces, was indeed poured into the baiter for each hundred hooks of the longline, so the “utilisation rate” were easily calculated from the observed mean of correctly baited hooks. About 50 % of unutilised fish chops could in turn be given once again to the baiter as they were still stiff enough.
As a rule all longlines were retrieved by both systems at a steady rate of 5-6 hooks per minute, independently from the operational depth and ground of each haul. Retrieving operations were also poorly influenced by bad weather, since in 7 hauls performed with comparatively stronger sea (3 after the Douglas visual scale of waves) on 130 - 350 m grounds we recorded the retrieving speed of 6.1+1.7 hooks/min. similar to the mean value 5.6+1.4 hooks/min. observed for all other experimental hauls (p> 0.30, Student test for unpaired groups of data, 46 d. f.).
The hanging ratios estimated for the 41 hauls carried out by the LLF1 longline showed that the gear formed volutes rising on average 1.0-3.5 ms from the sea floor. All hauls by system II and longlines LLF2 - LLF3 gave hanging ratios inferior to 1.00, which probably means that the gears were dragged on the fishing grounds. Lastly, hanging ratios estimated for the two commercial hauls seem imply that the fishing gear mainly formed 5 - 6 m vertical volutes.
b - Composition of catches and yields
During the tests of July-September 1993, 1,434 fishes of 26 species were caught for the total weight of 1,511.5 Kg. Four species - Squalus acanthias (L.), Conger conger (L.), Merluccius merluccius (L.) and Helicolenus
dactylopterus dactylopterus (Delaroche) - made up 73.6% in terms of
number and 76.3% in terms of weight of the total catches. Nine species were found only in the catch of one or two hauls or even as single individuals during the entire series of tests (Table 2).
In both commercial hauls a total of 565 ca fishes for an estimated total weight of 800 Kg were captured; among them the frequencies of specimens both of M. merluccius and Lepidopus caudatus (Euphrasen) were rather high, whilst catches of C. conger were low compared to the experimental trials (Table 3).
In Table 4 statistical descriptors of the length. and weight distributions of fishes caught at different depths during the series of experimental hauls are listed: as a rule the bathymetric range of the single species and size groups agree with data reported in the literature for the Southern Adriatic and other Mediterranean areas (BINI, 1968; BELLO and RIZZI, 1988; JARDAS, 1988; UNGARO et al., 1993). One exception is the capture of 43 large conger eels (average weight = 2.2+1.5 Kg) on comparatively shallow grounds (depth sector 100 - 150 m) of the fishing A area.
With concern to our experimental hauls, M. merluccius and C. conger made the bulk of the catches, with 372.3 and 380.5 Kg respectively. However, most individuals of both species were obtained in a small number of “lucky hauls”, so that half of the conger eel and hake catches came, considering only the fishing operations performed beyond 150 m, from two distinct groups of 5 hauls.
Similarly the catches of Selachians S. acanthias and Galeus melastomus Rafinesque were highly concentrated as 59 out 71 individuals of the former species fished in summer 1993 came from four hauls performed on 100 - l50 m grounds of the fishing A area (average number of baited hooks = 412+75), whilst 70 out of 105 individuals of the latter one were caught in a group of six hauls (average number of hooks baited = 367+66) carried out in the deepest sectors of the same area.
potential of the species composition of catches in discriminating groups of hauls similar to those defined on the basis of operational depths. Results in figure 3 show that hauls from the three depth sectors (100 - 150, 150 - 300, 300 - 400 m) are set apart, although groups are dispersed and scarcely homogeneous.
In figure 3 we see a similarity dendrogram obtained by cluster analysis (MOSSART and KAUFMAN, 1989) on data from 43 experimental hauls performed off the Apulian coast of the Southern Adriatic Sea (also this time 5 hauls were excluded for their poor yields); in the figure C. conger, M.
merluccius, S. elongata and H. d. dactylopterus are linked together, whilst
the catches of S. acanthias are connected with those of Trigla lucerna L. Other groups of fishes are Phycis phycis (L.), Scorpaena scrofa L. and Trigla
lyra L.; then Phycis blennoides (Brunnich), Polyprion americanus (Bloch &
Schneider) and Pagellus bogaraveo (Brunnich).
In order to find out several factors influencing the longlines’yields, all hauls performed in the fishing A area during summer 1993 were first depicted on nautical maps and then inscribed within the smallest possible circles defining distinct sub-areas. This procedure showed that “more profitable” and “less profitable” hauls (classified by comparison with the weighed average yield calculated on the entire set of 48 experimental hauls) were located in distinct subareas (named H1-H3and L1-L4respectively in Fig. 1).
If we put together data from all experimental hauls performed in L1-L4and in H1-H3, yields concerning C. conger, M. merluccius and the group “all fishes” are statistically different, both in the 150-300 and 300-400 depth sectors. As an example, weighed means and standard deviations of the whole catches result 8.2+1.8 vs. 3.6+2.9 Kg for different clusters of hauls carried out at depths lower than 300 m (p<0.0l Student’s test for unpaired groups of data with 14 d. f; see Table 5 for further details).
intermediate depth sector (Table 5). In the same way, no clear differences were observed on the faunistic composition of catches from sub-areas H1-H3 and L1-L4, although we did not perform formal tests about this.
The H3 sub-area was also selected by the fishermen of the mentioned longliner to perform both commercial hauls of summer and autumn 1995, registering high yields. Both hauls were carried out, exactly in the same site, on a 275 - 320 m deep ground, obtaining 14.1+6.4 fishes and 20.0+11.5 Kg for hundred baited hooks (Table 3).
d - Additional observations
In analysing the data from all experimental hauls of summer 1993, the number of conger eels caught during each haul resulted to increase with their average size. When all longlines were put together, an average rate of 2.27+1.54 conger-eels was observed in a group of nine hauls where these fishes weighed more than 1.8 Kg vs. the rate of 1.4+0.9 individuals registered in another group of 25 hauls where the average weight of the fishes ranged 0.4 and 1.6 Kg (p<0.05; Student’s test for unpaired groups of data, 32 d. f.).
Both LLF2 and LLF3 longlines fished better on P. bogaraveo than the LLF1 gear equipped with larger hooks. All hauls carried out in the fishing A area by the first two longlines gave, in the 150 - 300 m depth, an average catch of 2.0+0.9 individuals every hundred baited hooks, whereas the same rate resulted 0.5+0.1 for the fishing operations by LLF1 (p<0.01; Student’s test for unpaired groups of data, 9 d. f.).
On this matter we observe that the average body lengths of the blackspot breams caught during the above mentioned groups of experimental hauls were almost equal: 32.0+4.5 cm for the operations by LLF1 (N 18; hauls 7) vs. 32.0+4.0 cm for the specimens caught by LLF2-LLF3 (N = 28; hauls = 4). Moreover the conger-eels and seabreams caught during our trials were usually found with the hook positioned on the anterior edge of the mouth, instead of being swallowed as seen in both hauls by LLCM longline. Lastly, it is interesting to see that high catches of the Selachian Squalus
blainvillei (Risso) obtained in a set of 13 experimental hauls carried out
during summer 1992 near the Greek island of Cephalonia (NE Ionian Sea, area not shown in Fig. 1). These hauls were performed at depths from 135 to 440 meters baiting on average 393+110 hooks and obtaining 332 S.
Discussion
Longlining is supposed to be a selective fishing method, since larger fishes from few piscivorous species are usually taken (BJORDAL, 1990a and b; KENCHINGTON, 1996). Catches are therefore largely composed of mature individuals that spawned at least once before being fished, thus fish stocks should be exploitable more confidently and for longer time.
However longlining is a demanding fishing method, requiring large crews to ensure continuity to the work at sea (PEDERSEN, 1989; PIETRUCCI and ANTOLINI, 1990a); moreover, revenues largely depend on the numbers of hooks set every day as well as the prevalence of large specimens in the catch (B J O R D A L, 1990a; L O K K E B O R G and B J O R D A L, 1992; KENCHINGTON, 1996).
These problems could be alleviated by installing mechanisation systems to increase the numbers of hooks “worked” per unit of time. The fact that longlines of strong design have been used for long time along the Apulian coast of the Southern Adriatic Sea (PIETRUCCI and ANTOLINI, 1990a and b) would presumably ease, in absence of the EC fishing rule 1626/94, the spreading of such mechanisation systems in the area.
Analysis of our results is therefore of great relevance to evaluate fishing efficiency of the tested systems as well as their performance in making operations at sea faster.
I - Baiting and longline retrieval
All tests carried out during summers 1992 and 1993 point out that the baiting rates registered with S. pilchardus chops never exceeded 75.0% of the hooks fed to the baiter; when mackerel pieces were used instead, the rates of hooks correctly baited progressively increased with time to 90.0% ca. All these data agree with the manufacturers’directions (ANONYMOUS, 1984) and reports from independent observers (e. g. DAHM, 1986a and b) that hard baits, as Cephalopods or harengs, are needed to operate with the mentioned automation systems.
Automation systems performance during the retrieval of longlines could be hardly assessed, because fishermen needed to check the gears carefully before launching them again. Nevertheless checking operations seem to be partially accelerated in comparison with traditional procedures, since our unexperienced crew stored the longlines at an average rate of 450 snoods/hour/man similar to that registered by other authors on a specialised longliner (PIETRUCCI and ANTOLINI, 1990a).
II - longline positioning
In the Mediterranean longlines aiming at demersal fishes are usually positioned at some distance from the sea bed to minimise predation by epifaunal invertebrates as well as fortuitous entanglements (RUSSO, 1928; PIETRUCCI and ANTOLINI, 1990b). In additon these longlines are usually equipped with long monofilament snoods (up to 1.8 m each) that are supposed to catch more fish (SCACCINI et al., 1970; PIETRUCCI and ANTOLINI, 1990b).
Such design is a great obstacle to the mechanisation of fishing operations: this explains why automation systems like those we have tested are widespread in geographic areas where the existence of large rocky or sandy grounds on the continental shelf (e. g. off NE European coast, see Boillot et
al., 1971) makes it possible to use “simple” longlines to be directly set on the
sea floor (SKUD and HAMLEY, 1978; FLAGEUL, 1994). In fact the only longline supplied to us by the U.S. manufacturer of system I (LLF6 in Table 1) was composed of a mainline with an inner lead rope to make it steadily lay on the sea floor.
III - Composition of catches
Our results confirm that longlines are highly selective, since specimens from five species - G. melastomus, S. acanthias, C. conger, M. merluccius and H.
d. dactylopterus — numerically made up 81% of total fishes caught.
Moreover, in both commercial hauls of summer-autumn 1995 almost all fishes were M. merluccius, L. caudatus, S. elongata and H. d. dactylopterus. The species and size composition of catches clearly differed from that reported for trawling surveys carried out in the Southern Adriatic Sea at depths from 200 to 400 m (BELLO et al., 1988; UNGARO et al., 1993). In details, Scorpaenidae and the sparid P. bogaraveo made up about 15.0% , in terms of weight, of the marketable product obtained during our trials compared with 2.0 - 3.0% fractions registered in trawl catches.
Moreover S. elongata, fish presumed to be uncommon in the Southern Adriatic Sea (SCACCINI et al., 1970; BELLO and RIZZI, 1988), on the contrary appeared a somewhat relevant resource for longliners operating in the area.
Comparing our data with those reported on commercial longlining carried out some decades ago in different sub-areas of the Mediterranean Sea (PASQUINI, 1926; RUSSO, 1928; ARCIDIACONO, 1936; KIRINCIC and LEPETIC, 1955; SCACCINI, 1963; SCACCINI et al., 1970; BOLOGNARI
et al., 1971; DE ZIO et al., in press), no clear differences can be seen about
the species or size composition of the catches, due to the descriptive nature of most papers as well as to planopy of environmental and operative factors influencing yields and catch composition (SKUD and HAMLEY, 1978; LOKKEBORG et al. 1989; LOKKEBORG and BJORDAL, 1992; 1995; ENGAS and LOKKEBORG, 1994).
In Table 6 the size compositions of our catches are compared with data reported in the literature about the growth and length at maturity of different fishes (although often data are ascribed to populations from extra-mediterranenan areas) in order to make some guesses on the age structure of our samples and the fractions of mature individuals. In our samples the fractions of mature individuals presumably ranged from 9.1% for P.
bogarvaeo to ca 100.0% for M. merluccius and T. lucerna. Age composition
of our samples presumably ranged from 3 years or more for T. lyra individuals to a minimum age of 10 - 15 years for S. acanthias.
During our tests good fishing yields were also registered for S. acanthias, and G. melastomus, in summer 1993 and for S. blainvillei respectively one year earlier, so our longlines seemed to efficiently capture large and medium-sized Selachians. These results agree with the report by ERZINI et
al. (1998) that G. melastomus, Scyliorhinus canicula L. and Etmopterus pusillus (Lowe) numerically amounted to ca 35% of the catches obtained by
monofilament longlines at 200 - 700 m depths off the Algarve coast (southern Portugal).
IV - Fishing yields and species clusters
Our experimental data showed the relevance of the haul sites to determine fishing yields, since in the fishing A area distinct sub-areas of limited extension with higher and lower yields were recognised. Moreover the good catches registered in one of the sub-areas during both experimental hauls of summer 1993 and the commercial ones performed two years later imply that more productive sectors preserve, to some extent, such productivity over time.
These observations agree with many reports in the literature about positive effects on the fishing yields of persistent factors such as sharp slopes (LEHODEY et al., 1994), local currents (LOKKEBORG et al., 1989; ENGAS and LOKKEBORG, 1994) and untrawlable grounds (SCACCINI, 1963; RUSSELL et al., 1988).
(Tables 3 and 5).
Our data clearly showed that the rather large hooks used for the LLFl, LLF2 and LLF3 longlines reduced the catches of small-mouth fishes such as C.
conger and P. bogaraveo. In this view the overlapping body sizes registered
for blackspot breams caught by different experimental longlines are explained assuming that all hooks were so large that only the biggest individuals could be captured. These results agree with the report by SOUSA
et al. (1999) saying that in comparative tests performed in the Azores on the
fishing efficiency of distinct hooks best yields for P. bogaraveo were registered using hooks measuring 23-8-7 mm ca.
Bigger baits were probably another important factor determining the higher yields seen in the couple of commercial hauls; the great care paid by the fishermen to select large pilchards (otherwise two fishes were put together on the hook, our personal observation) clearly showed the relevance attached to the “bait factor”. This agrees with the general notion that the greater preys increase the average size of the predator fishes caught at sea (P I T C H E R and H A RT, 1985; L O K K E B O R G and B J O R D A L, 1992 and 1995). Nevertheless a reduction of bait costs is an important goal for commercial longliners (L O K K E B O R G and B J O R D A L, 1995), so small chops were used during our trials.
Higher volutes of LLCM (as inferred on the basis of the longline’s theoretical hanging ratios) may be the third factor determining the higher yields of L. caudatus and M. merluccius registered for this gear since partly lifted bottom longlines are often used to catch Gadoids or frost-fishes (P O L I C E, 1919; FA O, 1987; D E M E S T R E et al., 1993; K E N C H I N G TO N, 1996).
Some clues on the nature of the fishing grounds were our experimental hauls were paerfomed can be drawn from the species clusters depicted in Fig. 3: combined catches of Phycis phycis (L.), Scorpaena scrofa and Trigla lyra (L.), all fishes often located in medium-depth rocky areas (BINI, 1968), may imply that longlines operated on these areas or neighbouring sectors; the cluster P. blennoides, P. americanus and P. bogaraveo, presumably reflects catches from medium-depth soft or mixed areas as the above mentioned species are preferentially located on these grounds (BINI, 1970; ORSI-RELINI and FIDA, 1992). In our opinion, the strong correlation between S.
acanthias and T. lucerna was due instead to the fact that both species are
KIRNOSOVA, 1989). Analogously, the positive correlation found between the catches of C. conger, M. merluccius, S. elongata and H. d. dactylopterus presumably derives from wide depth distributions observed for these species (Table 4).
Conclusion
Both tested automation systems turned out to be of sturdy structure and sound mechanics since we could operate down to 400 m without registering failures.
Baiting tests by random system I registered fair “success rate”, even when soft-flesh baits as pilchards were used. Moreover, little extra consumption of baits was requested in comparison with the traditional hand-made baiting method.
The fishing yields registered with the LLF1 longline fell far behind those registered for a couple of commercial hauls performed off Mola di Bari in summer-autumn 1995. Such lower yields were presumably due to the simpler rigging adopted for the experimental longlines which made them compatible with the automation system but less efficient in catching fishes. Nevertheless, the yields observed for these longlines turned to be economically viable (Dr. M. FERRETTI, pers. comm.*) when they were linearly extended to the average number of number of baited hooks — 1,500 hooks per man a a day - that experienced crews are expected to set daily on the basis of all data available from distinct geographic areas (CADE, 1981; ANONYMOUS, 1984).
Our data showed that bottom longlining could adversely affect the consistence of the Selachian stocks as these fishes get late to maturity and are voracious predators often eating the baits. Similarly the comparatively high yields registered for the couple of hauls performed by commercial longlines induce to suspect that the catches obtained by longliners operating in the Southern Adriatic Sea (ca 15 boats in 1998) could adversely influence the spawner stocks of Teleosts as M. merluccius and T. lucerna whose populations are heavily exploited by local trawlers.
Acknowledgements
and to A. Petrucci of IRPEM-CNR (Ancona, Italy) and to C. Papacostantinou of NCMR (Athens, Greece) for the help offered in various steps of this study.
REFERENCES
ANONYMOUS, 1984 - An Atlantic Fisherman’s Guide to Longlining systems. Dept. of Fisheries and Oceans Directorate Communications, Ottawa, 47 pp. (cod. Fs 23-63/E)
ARCIDIACONOF., 1936 - Per una maggiore conoscenza della pesca con consi
nel golfo di Squillace. Boll. Pesca Piscic. Idrobiol., XII, (1): 35 -50
ARMITAGEP., 1977 - Statistica medica. II˚ ediz., Feltrinelli, Milano, 496 pp. BARONJ., 1985a - Les Triglides (Teleosteens, Scorpaeniformes) de la baie de
Douarnez. I La croissance de Eutrigla gurnardus, Trigla lucerna, Trigloporus
lastoviza et Aspitrigla cuculus. Cybium 3eme série, 9:127 - 144
BARONJ., 1985b - Les Triglides (Teleosteens, Scorpaeniformes) de la baie de
Douarnez. II La re p roduction de Eutrigla gurnardus, Trigla lucerna,
Trigloporus lastoviza et Aspitrigla cuculus. Cybium 3eme série, 9: 255 - 281 BELLO G., MARANO G., RIZZI E., 1988 - Risorse demersali del Basso
Adriatico. Risultati del primo anno di indagine. Atti Seminario
C.N.R-M.M.M , 10 - 11 novembre 1986, Roma: 1531 - 1556
BELLO G., RIZZI E., 1988 - I Teleostei raccolti nell’Adriatico meridionale
nelle campagne sperimentali di pesca a strascico 1985-87. Quad. Ist. Ric.
Pesca Marittima, 5, 1: 77 - 90
BINI G. 1968 - Atlante dei pesci delle coste italiane. Mondo Sommerso, Milano, I - IX
BJORDAL A., 1990a - Recent developments in longline fishing - catching
performance and conservation aspects. In: Voigtm. N., Boita J. R: (Eds) Advances in Fisheries Technology and Biotechnology for Incre a s e d Profitability. Technomic, Lancaster (PA), USA: 19 - 24
BJORDALA., 1990b - Application of research in fish harvesting by passive
gears in Norway, particularly longlining. In: Voigt M. N., Botta J. R: (Eds) Advances in Fisheries Technology and Biotechnology for Incre a s e d Profitability. Technomic, Lancaster (PA), USA: 23 - 32
BOLOGNARI A., BUTA G., CAVALLARO G., 1971 - Risultati delle pescate
della Sicilia Orientale. Boll. Pesca Piscic. Idrobiol. XXVI, 1-2: 21 - 46
BOILLOT G., BOUYSSE P., LAMBOY M., 1971 - Morphology, sediments and
Quaternary history of the continental shelf between the Straits of Dover and Cape Finisterre. In: DELANEY F.M. (ed.) The Geology of the East Atlantic
Margin, 3. Europe IGS Rep., N˚ 70/15 : 79 - 90
CADER, 1981 - Catch’81 Supplement. New Zealand Ministry of Agriculture and Fisheries Media Service , February 1981: 1 - 20
CARTES J. E., COMPANY J. B., MAYNOU F., 1994 - Deep-water decapod
crustacean communities in the Northwestern Mediterranean: influence of submarine canyons and season. Marine Biol, 120: 221 - 229
CAU A., MANCONI P., 1983 - Sex-ratio and spatial displacement in Conger conger (L.) Rapp. Comm. int. Mer Médit., 28, 5: 93 - 96
CAUA., MANCONI P., 1984 - Relationship of feeding reproductive cycle and
bathymetric distribution in Conger conger. Marine Biol., 81: 147 - 151
CHOPINF., 1994 - The battle of the lights: as Japan’s squid jiggers face up
fierce competition. Fishing News Intern., January 1994: 8 - 11
CS A N A D YG. T., CH U R C H I L LJ. H., BU T M A NB., 1988 - Near bottom currents over
the continental slope in the Mid-Bight. Continental Shelf Res., 8: 653 - 672
DA H M E., 1986a - Langleinenversuche in der Ostsee. Fischerblatt., 34: 236 - 239
DAHME., 1986b - Erster Einsatz eines automatischen Langleinensystems auf
der "Solea”. Fischerblatt., 34: 239 - 242
DE ZIO V., UNGARO N., VLORAA., STRIPPOLI G., 1996 - Analisi strutturale
dello stock di nasello del Basso Adriatico sfruttato dalla pesca con palangari di fondo. Testo proposto al Seminario S.I.B.M. "Risorse demersali dei Mari
Italiani”, Fano, 20 - 22 Marzo 1996 (Quotation authorised)
D’ONGHIAG., MASTROTAROF., PANZA M., 1996 - On the growth of rockfish Helicolenus dactylopterus (Delaroche, 1809) from the Ionian Sea. FAO Fish. Rep., 533 (suppl.): 143 - 152
ENGASA., LOKKEBORGS., 1994 - Abundance estimation using bottom gillnet
and longline. The role of fish behaviour. In Ferno A., S. Olsen (Eds) - Marine fish behaviour in capture and abundance estimation. Fishing News Books,
Oxford: 134 - 165
ERZINIK., GONCALVESJ. M S., BENTESL. , LINOP. G., RIBEIROJ., 1998 - The
Sept 1998, l2 pp
FA B B R I A., GA L L I G N A N I P., 1972 - R i c e rche geomorfologiche e
sedimentologiche nell’Adriatico meridionale. Giornale di Geologia,
XXXVIII: 453 - 498
FAHYE., 1989 - The spurdog Squalus acanthias (L.) fishery in the south west
Ireland. Ir. Fish. Invest. Ser. B , 32: 1 - 22
FANNONE., FAHYE., REILLY R O., 1990 - Maturation in female conger eel, Conger conger (L.). J. Fish Biol., 36: 275 - 276
FAO, 1987 - Catalogue of small-scale fishing gear. 2nd edition, Fishing News Books, Farnham, 224 pp.
FAO, 1993 - Review of the state of world marine fishery resources. FAO Techn. Pap., 335: 1 - 136
FLAGEUL L., 1994 - La palangre en Manche-Ouest. France Eco-Peche, Fevrier 1994: 50 - 51
FLAMIGNI C., 1984 - Preliminary utilization of trawl survey data for hake (Merluccius merluccius L.) population dynamics in the Adriatic Sea. FAO Fish. Rep., 290:109 - 115
GALLIGNANIP., 1973 - I sedimenti della piattaforma continentale pugliese da
Bari a Torre Canne. Giornale di Geologia, XXXIX: 115 - 131
GRUBISIC F., 1959 - Changes in fishing methods and gear on the Eastern
Adriatic coast. Proc. gen. Fish Coun. Medit. 5: 351 - 353
JARDASI., 1988 - Distribution of the Adriatic fishes of the Triglidae family
as affected by ecological factors. FAO Fish. Rep., 349: 147 - 151
KENCHINGTONT. J., 1996 - Long-term stability and change in the commercial
groundfish longline fishing grounds of the northwest Atlantic. Fish. Res., 25:
139 - 154
KETCHEN K. S., 1972 - Size at maturity, fecundity and embryonic growth of
the spiny dogfish (Squalus acanthias) in British Columbia waters. J. Fish.
Res. Board Canada , 29: 1717 - 1723
KIRINCIC J., LEPETIC V., 1955 - Recherches sur l’ichthyobenthos dans les
profondeurs de l’Adriatique meridionale et possibilitè d’explotation au moyen des palangres. Acta Adriatica VII: 1 - 109
KIRNOSOVA I. P., 1989 - Reproductive biology of spiny dogfish Squalus acanthias in the Black Sea. Journ. Ichthyol., 29, 3: 21 - 26
KRUG M H., 1990 - The Azorean blackspot seabream, Pagellus bogaraveo
LEHODEYP., MARCHALP., GRANDPERRINR, 1994 - Modelling the distribution
of alfonsino, Beryx splendens, over the seamounts of New Caledonia. Fish.
Bull., 92: 748 - 759
LOKKEBORG S., BJORDAL A., FERNO A., 1989 - Responses of cod (Gadus morhua) and haddock (Melanogrammus aeglefinus) to baited hooks in the
natural environment. Can. J. Fish. Aquat. Sci., 46: 1478 - 1483
LOKKEBORGS., BJORDALA., 1992 - Species and size selectivity in longline
fishing: a review. Fish. Res. 13: 311 - 322
LOKKEBORG S., BJORDALA., 1995 - Size-selective effects of increasing bait
size using an inedible body on longline hooks. Fish. Res. 24: 273 - 280
MARANO G., ROSITANIL., UNGARO N., DE ZIO V., 1988 - Cours des prises
avec palangre de surface dans l’Adriatique du Sud (còtes italiennes), triennat 1984-86. FAO Fish. Rep., 349: 112 - 120
MASSART D. L., KAUFMAN L., 1989 - The interpretation of analytical
chemical data by the use of cluster analysis. R E. Krieger Publ. Co., Malabar
(FL), USA, 237 pp.
MILLER J. C., MILLERJ. N., 1988 - Statistics for analytical chemistry. Ellis Horwood Limited, Chichester, U.K., 227 pp.
MURPHYG. I., 1960 - Estimating abundance from longline catches. J. Fish. Res. Bd. Canada 17: 33 - 40
ORSI - RELINIL., FIDAB., 1992 - Note di biologia di Pagellus bogaraveo in
Mar Ligure. Oebalia, Suppl. XVII: 81 - 86
PAPACOSTANTINOU K., 1981 - Age and growth of piper, Trigla lyra, in
Saronikos Gulf (Greece). Cybium 3e série, 5: 73 - 87
PAPACOSTANTINOUK., PETRAKISG., CARAGITSOUE., MYTILINEOUC., 1992
-Preliminary study on the biology of piper (Trigla lyra L., 1758) in the Aegean Sea. FAO Fish. Rep., 477: 127 - 137
PASQUINI P., 1926 - Per una maggiore conoscenza della pesca adriatica ed
insulare. Boll. Pesca Piscic. Idrobiol., II, 2: 3 - 64
PE D E R S E N F., 1989 - The economic and environmental benefits of
mechanised longlining. World Fishing, Feb. 1989: 10 - 11
PEIRANO A., TUNESI L., 1986 - Preliminary notes on the biology of Helicolenus d. dactylopterus (Delaroche) in the Ligurian Sea. Rapp. Comm. int. Mer Médit., 30, 2: 233
PIETRUCCIA., ANTOLINIB., 1990 - Rilancio della pesca con il palangaro di
fondo. Gazzettino Pesca , Aprile 1990: 32 - 38
palangaro di fondo. Gazzettino Pesca, Novembre 1990: 15 - 20
PITCHER T. J., HART P. J. B., 1985 - Fisheries ecology. Croom Helm, Beckenham, 414 pp.
POLICE G., 1919 - Per la pesca a profondità. Modificazione all’arnese da
pesca. Boll. Soc. Nat. Napoli, XXXII: 92 - 100
RAGONESE S., REALE B., 1995 - Distribuzione e crescita dello scorfano di
fondale, Helicolenus dactylopterus dactylopterus (Delaroche, 1809), nello Stretto di Sicilia (Mar Mediterraneo). Biol. Mar. Medit. 2, 2: 269 - 273
RICKERW. E., 1975 - Computation and interpretation of biological statistics
of fish populations. Bull. Fish. Res. Board. Can. 191: 1 - 382
ROMANELLI M., PALLADINO S., TARULLI E., FERRETTI M., 1996 - Stima
dell’accrescimento di Helicolenus d. dactylopterus (Delaroche) in Adriatico Meridionale tramite esame delle sagittae di esemplari prelevati con reti a strascico e palangari di fondo. Biol. Mar. Medit., 4, 1: 554 - 556
RUSSEL M, GUTHERZ E. J., BARANS C. H., 1988 - Evaluation of demersal
longline gear off South Carolina and Puerto Rico with emphasis on deep-water reef fish stock. Mar. Fish. Rev. 50: 26 - 31
RUSSOA., 1928 - Studi sulla pesca nel golfo di Catania (parte quarta). Boll. Pesca Piscic. Idrobiol., IV, 5: 495 - 543
SCACCINI A., 1963 - Résultats des recherches sur l’exploitation des eaux
Tyrrhéniennes de la Calabre (13 juin-14 juillet 1962). Proc. Gen. Fish Coun.
Medit. 7: 75 - 83
SCACCINI A., PICCINETTI C., SARÀ R, 1970 - Stato attuale della pesca in
acque profonde nei mari italiani. Boll. Pesca Piscic. Idrobiol., XXV, 1: 8
-35
SKUD B. E., HAMLEY J. M., 1978 - Factors affecting longline catch and
effort. Int. Pac. Halib. Commn. Sci. Rep., 64:1 - 50
SOUSA F., ISIDROE., ERZINIK., 1999 - Semi-pelagic longline selectivity for
two demersal species from the Azores: the black spot sea bream (Pagellus
b o g a r a v e o ) and the bluemouth rockfish (Helicolenus dactylopterus dactylopterus). Fish. Res., 41: 25 - 35
TORTONESE E., 1956 - Leptocardia, Ciclostomata Selachii. Fauna d’Italia, Ed. Calderini, Bologna, II, 334 pp.
UNGARO N., RIZZI E., MARANO G., 1993 - Nota sulla biologia e pesca di Merluccius merluccius (L.) nell’Adriatico pugliese. Biol. Mar. Medit., 2, 1: 32 - 334
methods of the world. Fishing News Books, Farnham, Surrey (UK), 80 - 105
ZUPANOVICS., 1968 - Study of hake (Merluccius merluccius L.) biology and
TABLE4 - Synopsis of the average sizes and weights recorded for different fishes during our
Fig. 1 - Map of the three fishing areas where experimental longlines were tested and
locations of the “scarcely productive” (L) and “highly productive” (H) sites found in the fishing A area.
Fig. 2 - Discriminant analysis based on the species composisiton of the catches obtained in
Fig. 3 - Dendrogram of the fish clusters obtained on the basis of the species composition of
